trainer_state.json

{
  "best_metric": 0.9237333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0706/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.630785584449768,
      "learning_rate": 0.00042857142857142855,
      "loss": 0.72,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8298666666666666,
      "eval_loss": 0.5990007519721985,
      "eval_runtime": 4.0876,
      "eval_samples_per_second": 917.417,
      "eval_steps_per_second": 14.434,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2649858891963959,
      "learning_rate": 0.00035714285714285714,
      "loss": 0.3462,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.848,
      "eval_loss": 0.5488991737365723,
      "eval_runtime": 4.0912,
      "eval_samples_per_second": 916.611,
      "eval_steps_per_second": 14.421,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4727197289466858,
      "learning_rate": 0.0002857142857142857,
      "loss": 0.197,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8850666666666667,
      "eval_loss": 0.42757630348205566,
      "eval_runtime": 4.3598,
      "eval_samples_per_second": 860.131,
      "eval_steps_per_second": 13.533,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.0665178894996643,
      "learning_rate": 0.00021428571428571427,
      "loss": 0.0997,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8909333333333334,
      "eval_loss": 0.40692251920700073,
      "eval_runtime": 4.1233,
      "eval_samples_per_second": 909.465,
      "eval_steps_per_second": 14.309,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.004460092633962631,
      "learning_rate": 0.00014285714285714284,
      "loss": 0.0456,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.916,
      "eval_loss": 0.33850276470184326,
      "eval_runtime": 4.3743,
      "eval_samples_per_second": 857.271,
      "eval_steps_per_second": 13.488,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.05313340201973915,
      "learning_rate": 7.142857142857142e-05,
      "loss": 0.0123,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9221333333333334,
      "eval_loss": 0.31963443756103516,
      "eval_runtime": 4.359,
      "eval_samples_per_second": 860.287,
      "eval_steps_per_second": 13.535,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.1450021450021452e-07,
      "loss": 0.0022,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9237333333333333,
      "eval_loss": 0.308350533246994,
      "eval_runtime": 4.348,
      "eval_samples_per_second": 862.465,
      "eval_steps_per_second": 13.569,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}